- Home
- Skills
- Technical Documentation
- Evidence-Bound Engineering Handoff
Works with the AI Agents you already use
Evidence-Bound Engineering Handoff
Prepare an engineering handoff that connects scoped changes to exact source versions, test evidence, unresolved failures and the next owner decision.
$5
Evidence-Bound Engineering Handoff
Example session with this skill installed
Synthetic local fixture: prepare an actual-results transfer for changing a title. The title must be present and fit eight characters. Review only; no implementation or dispatch.
UTF-8 files, each ending with one LF:
- baseline/title.txt: Demo
- candidate/title.txt: Summer Demo
- baseline/unrelated.txt: blue
- candidate/unrelated.txt: green
Use the supplied synthetic check evidence: title present PASS; eight-character limit FAIL (11 characters); natural GUI interaction NOT_RUN. Keep owner acceptance pending.
- Read your context and instructions
- Compiled the evidence-bound engineering handoff
Worked example: synthetic title change
This is a synthetic local exercise, not a customer project or a demand signal.
The test was executed on October 23, 2026. The following is the handoff produced
by applying this pilot's instructions to that fixture. No separate reviewer or
second agent evaluated the output.
Given brief and raw inputs
Prepare an actual-results transfer for changing a title. The title must be
present and fit eight characters. Review only; no implementation or dispatch.
The UTF-8 fixture files have these exact contents, with one LF at the end:
| File | Contents represented as a string |
|---|---|
| baseline/title.txt | "Demo\n" |
| candidate/title.txt | "Summer Demo\n" |
| baseline/unrelated.txt | "blue\n" |
| candidate/unrelated.txt | "green\n" |
Actual-results transfer
Outcome and scope: The candidate changes title.txt from Demo to
Summer Demo. It implements the requested title change but fails the fixture's
eight-character acceptance criterion. An unrelated notes change from blue to
green is visible and excluded from the task's claimed delta. There are no new
or removed scoped files. No original fixture file was modified during checks.
Source identity: This fixture is not a Git checkout; no HEAD is claimed.
SHA-256 binds the baseline title.txt to
000fa753b5b9531c958e09bfce1fc74bbbddb72de3bc2a7ef5dfc67c467cbae3
and final title.txt to
d8e30dcb6f8e1a369f9905d7ff5bdc4828f8825894a049aed4cc42167b30ec91.
Executed checks: Title present: PASS. Title length: FAIL, 11 characters
against a maximum of eight. The local JSON result records both outcomes.
Natural GUI interaction: NOT RUN because the fixture contains no application.
The JSON result SHA-256 is
229ba724e14c91cfbfadc6e3a95bfb08b365f95c14b3a5dad4791c2b6a1a96c4.
All evidence is synthetic; there is no usability or production-performance claim.
Evidence gaps: A separate stale-title scenario differs from the final hash
and cannot replace the recorded candidate. A proposed screenshot is absent;
visual fit remains unverified. The failed length check remains a failure.
Review boundary: A reviewer can recount the final title and compare its
hash with the recorded result. No reviewer has been appointed for this fixture.
Owner acceptance is PENDING. The next decision is whether to authorize a shorter
title or revise the criterion. This handoff does not authorize that correction.
Connects securely to your tools. The creator never sees your data.
About this skill
Prepare an engineering handoff that connects scoped changes to exact source versions, test evidence, unresolved failures and the next owner decision.
Use it when transferring a task or preparing work for review. It guides your agent to distinguish task changes from unrelated edits, preserve passed, failed and unrun checks, identify stale evidence, and keep technical results separate from owner acceptance. Includes self-contained instructions and a worked example.
Requires an agent with access to your approved brief, relevant files and suitable read/hash tools. Tested in one fresh Codex agent on a synthetic task and a small continuation. Other agents, live integrations and production use are unverified. No time-saving, sales or correctness guarantee is claimed.
Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for Technical Documentation.
Test evidence is limited to synthetic exercises, including one fresh Codex agent and a small continuation. Other agents, live integrations, production reliability, measured time savings and buyer demand are unverified. Hashes establish byte identity, not correctness. Preparing a handoff does not authorize implementation, commits, pushes, publication or owner acceptance.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before buying a skill.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Money-back guarantee
Doesn't do what its page promises? You get your money back.
Trust & safety
Security scan
Security scan passed 24 days ago
Requires
Nothing to install.
Permissions
Reads user-approved task instructions, baseline/source files and test evidence to prepare a handoff. Writes a report only when requested. No external service, credential or environment-variable access is required.